Customer segmentation

Customer segmentation is the practice of splitting your existing customers into groups that share traits such as age, location or buying habits, so each group gets its own message and offer.

What is customer segmentation?

TechTarget defines customer segmentation as dividing a customer base into groups of people with similar characteristics relevant to marketing, such as age, gender, interests and spending habits. The four most common types of data used to segment are demographic, geographic, psychographic and behavioral (TechTarget, What is customer segmentation, accessed October 2026).

The data comes from your first-party data in a CRM or CDP: order history, purchase channel, responses to messages.

RFM segmentation

RFM scores each customer on three numbers taken from their transaction history:

• Recency (R): how long since the last purchase, usually counted in days.
• Frequency (F): number of orders in the period, for example 12 months.
• Monetary (M): total amount the customer spent in the period.

Each number is split into 5 groups and scored from 1 to 5, with 5 being best. The three scores are combined into a code such as 555 or 151. All you need is a customer ID, an order date and an order value (rfm package for R, RFM Introduction, October 2026).

TechTarget calls the 5-5-5 group your biggest customers. The 5-1-X group (bought recently, only once) are new customers. The 1-X-5 group used to spend a lot but has not come back in a long time (TechTarget, RFM analysis, accessed October 2026).

RFM example for a cosmetics shop (hypothetical)

An online cosmetics shop has 10,000 customers who bought in the past 12 months. After RFM scoring, three groups stand out:

• 500 customers in 5-5-5: bought in the last 30 days, 6 or more orders, spent over VND 4 million.
• 1,200 customers in 5-1-X: made their first purchase this month.
• 800 customers in 1-X-5: spent over VND 4 million but have not bought in more than 6 months.

Sending a VND 100,000 voucher to all 10,000 customers costs VND 200 million if 20% use it, and part of that is a discount for the 500 customers in 5-5-5 who would have bought again anyway.

The segmented approach: only the 800 customers in 1-X-5 get the voucher, by ZNS (Zalo’s business messaging service) or email. If 10% (80 customers) come back with an average order of VND 600,000, the shop takes in VND 48 million and spends VND 8 million on vouchers. The 5-5-5 group gets new product news instead of a discount and is a natural fit for the top tier of a loyalty program. New customers get product how-to content to earn a second order.

Using segments in advertising

To run Google ads to a segment, you upload the list to Customer Match. List memberships expire after 540 days. A list needs at least 100 members added or updated within that window to stay eligible, so Google recommends continuous CRM sync (Google Ads Help, About Customer Match, accessed October 2026).

Names, phone numbers and purchase history are personal data. Using them to segment and advertise in Vietnam has to follow the Personal Data Protection Law No. 91/2025/QH15 (see our guide to Vietnam’s personal data protection law for marketers).

Customer segmentation vs related terms

Frequently asked questions

Does a small shop need customer segmentation?

Yes, once you have a few hundred customers with order history. RFM works in a spreadsheet. Without purchase history, start by segmenting on traffic source and the products people ask about.

Can an app that does not sell anything use RFM?

Yes, through a variant called RFE: replace Monetary with an engagement measure such as views or time spent (TechTarget).

If you want to split ad budgets and follow-up messages by customer group, see Adtek’s Brand Commerce page.